Like all three-dimensional objects, the limb consists of three axes. The long-term goal of this work is to understand the interactions of various signaling and genetic pathways that pattern these three axes of the developing limb of the mouse embryo. This will not only tell us how a limb forms in utero, but will tell us how certain signaling/genetic pathways interact with each other in other normal contexts in the developing embryo as well as in abnormal contexts such as cancer and congenital disease. The developing limb is ideal for such studies because a wealth of genetic and experimental embryological data provides a rich context in which to place new data. Previously we have shown an essential role for fibroblast growth factor FGF signaling during limb development by examining mice lacking genes that encode FGF ligands (Lewandoski et al. 2000 Nature Genetics 28:167, Sun et al 2000 Nature Genetics 25: 6 ). However, the complexity caused by Fgf gene redundancy has led us to consider approaching the problem by examining FGF receptor mutants leading to the insight that FGF signaling controls limb size (Verheyden JM et al 2005Development 132:4235). We are also studying the inter-relationship between FGF and BMP signaling. Current models of limb development suggest that BMP signaling plays a role in controlling the normal cell death that occurs in mesenchymal interdigit cells, sculpting the final digit pattern, but the exact role of BMPs in this process are unknown. By simultaneously inactivating the Bmp receptor gene, Bmpr1a as well as Fgf8 and Fgf4 specifically in the limb bud ectoderm, we have produced genetic evidence for a novel model in which the surface ectoderm must receive a BMP signal, resulting in down regulation of Fgfs which in turn induces apoptosis of the underlying mesenchyme. Thus both BMPs and FGF cooperate to control the essential cell function of programmed cell death. We are extending these studies by studying the role of BMP and FGF signaling in various aspects of limb development using mouse lines that express Cre in specific region of the developing limb. For example the only way to test the hypothesis that BMPs act as direct effectors of cell death is to inactivate BMPs receptors only in the lineage that undergoes cells death, without affect genes in nearby cells. We have achieved this using new Cre lines which allow Cre-mediated gene inactivation in these lineages. With these lines are asking: are BMPs are direct effectors of normal programmed cell death? If not, how is programmed cell death controlled? If so, how do BMPs achieve this endpont?